- From: Philip TAYLOR [PC87S-O/XP] <P.Taylor@Rhul.Ac.Uk>
- Date: Mon, 28 Jul 2003 16:16:58 +0100
- To: lholst@robotics.lu.se
- CC: Olivier Thereaux <ot@w3.org>, www-validator@w3.org
Lars Holst wrote: > [snip] > >Not sure about this either. If someone tries to validate a file, the > >validator should probably validate this file, not whatever it refreshes > >to after n seconds (I know, the case n=0 is a limit case but better > >forget this practice anyway). > > Why? The redirect is there for a reason, right? I really don't see why it > shouldn't follow it. OK, you have a refresh with period zero (which doesn't work reliably, so a period of one second would be better), and you would like the validator to validate the page /after/ the refresh/redirect has taken place. Supposing that the page to which you redirect has a further refresh/redirect, this time with a delay of two seconds; would you like the validator to validate /that/ page rather than either of the first two ? And if that page too has a refresh/redirect with an even greater delay, would you like the validator to follow that one as well ? If so, what is your cut-off point ? How long should the delay be in a refresh/redirect before you don't expect the validator to follow it ? > I do appreciate your pointer, but I'd be even more grateful for a link to a > page that actually explains the HTTP-based feature that I should use. You will need to be aware of which server you are using; for at last three common server, there was already a link to the documentation for that server in the original answer to which you refer; if you are using a less-common server, you will need to consult the server documentation. For a simple overview of what your server will need to accomplish (but with no guidance as to how to achieve it), see http://www.w3.org/Protocols/rfc2616/rfc2616-sec10.html#sec10.3 Philip Taylor, RHBNC
Received on Monday, 28 July 2003 11:19:12 UTC